On October 22, at 18:15, Maria Becker will give a public talk at the Interdisciplinary Forum of Digital Textual Sciences (InFoDiTex) with the title “Detecting Controversial Concepts within Discourses Using Quantitative Methods”.
We investigate a corpus of 72.000 German newspaper articles on climate change by applying quantitative methods such as keyword analysis or semantic tagging in order to find controversial topics and concepts. Our approach is divided into four steps with varying granularity: (a) Detecting Central Topics, (b) Detecting Central Subtopics, (c) Detecting Controversial Topics & Concepts and (d) Detecting Central Controversial Texts. Our contributions are twofold: (1) We reveal interesting insights into central controversial topics, actors, foci and points of view and styles of discussion and (2) we suggest a method for detecting Controversial Concepts within Textual Corpora using quantitative methods which we consider to be applicable to any other kind of discourse.
For more information, see here.